Global attractivity of the recursive sequence \(x_{n+1}=(\alpha-\beta x_{n-1})/(\gamma+x_{n})\) (Q1827026)

The authors consider the following rational recursive sequence and investigate its global attractivity \[ (1) \quad x_{n+1}=(\alpha - \beta x_{n-1})/(\gamma +x_n) \qquad\text{where}\qquad (2)\quad \alpha, \beta, \gamma > 0, \quad \text{and} \quad \gamma > \beta. \] Under condition (2) Equ.(1) has two equilibrium points \[ \overline{x}_i=\frac{-(\beta+\gamma) \pm \sqrt{(\beta+\gamma)^2+4\alpha}}{2},\quad i=1,2. \] They prove the following theorems: Theorem: The positive equilibrium point \(\overline {x}_1\) is locally asymptotically stable, while the negative equilibrium point \(\overline{x}_2\) is unstable (saddle point). Theorem: If there exists \(k \geq 2\) such that \(\gamma \geq k \alpha/\beta\), and \(\alpha \geq k \beta^2\) hold, then the positive equilibrium point \(\overline{x}_1\) of Eq.(1) is a global attractor with basin \[ S=[-(k-1)\alpha/\beta,\alpha/\beta]^2 \bigcup_{\alpha/\beta \leq M \leq (2k+1)\alpha/\beta} \left\{ M \right\} \times [M-2k\alpha/\beta,\alpha/\beta]. \] The definitions of asymptotically stable, equilibrium point and basin are standard ones. The method used here is an application of the basic criterion of \textit{M. R. S. Kulenovic, G. Ladas} and \textit{N. R. Prokp} [Comput. Math. Appl. 41, No. 5--6, 671--678 (2001; Zbl 0985.39017)] for the equilibrium point of the linearized equation associated with the Eq. (1).